# Reader tools From the author of Intelligence Capitalism (https://corelooppress.com/titles/intelligence-capitalism.html). HTML version: https://corelooppress.com/tools/ Twelve short, free tools in two groups. Tools 01 to 03 reflect your own answers about AI and work; they are conversation starters, not scientific surveys. Tools 04 to 11 use published research, official statistics, filings or open datasets. Tool 12 teaches source-backed distinctions among model access, rights, disclosure and practical use. Each evidence view identifies its source, date and limit. Tool 09, [Frontier IC25](https://corelooppress.com/tools/say-versus-spend/frontier-ic25/), lives on its own page and accompanies tool 08. It follows 25 model builders, scale enablers and AI-native deployers through linked evidence and one dated, model-level Epoch Capabilities Index view. Profiles put a source-linked fact or milestone first, retain detailed limits under an expandable panel, and provide unique share URLs plus landscape and portrait PNG cards. It is an editorial role map, not an extension of the IC 100 public-company cohort or a universal model leaderboard. Use the work tools for self-reflection (01 to 03), occupational or business evidence (04 to 05), compute concentration (06), electricity demand over time (07), company filing language (08), frontier organizations and models (09), facility resources (10), national electricity-system context (11), or model-release terminology (12). These subjects are complementary; a national electricity total does not answer a facility-connection question, and a company profile does not establish one openness status for all its models. ## About you and your work ### 01. Where do you stand on AI? https://corelooppress.com/tools/where-do-i-stand/ React to ten statements about AI, then see where your own answers place you on a map of views. Answers stay in the browser and are not stored. ### 02. What would you hate to lose? https://corelooppress.com/tools/what-would-you-hate-to-lose/ Choose three things you would hate to lose to AI, then see your picks beside the live tally on this site. Picks and country are stored without names or emails; the worldwide tally appears once at least 10 people have taken part, and a country tally once that country has at least 10 responses. Descriptive only. ### 03. What protects your work from AI? https://corelooppress.com/tools/exposure-audit/ Answer seven questions about your work, then see a profile of exposure and protection built from your answers. Scores are a reflection of the reader's answers and fixed editorial starting values, not measured or published data. Answers stay in the browser. An optional email sign-up for the next tool is offered after the result. Replaces the Occupational Exposure Audit (/audit-lite/, which now redirects here). ## About the published data ### 04. Is your job exposed to AI? https://corelooppress.com/tools/your-job-and-ai/ Guess how exposed a job is to AI, then see the research estimate beside observed use, jobs and pay. Sources: Eloundou et al. (2023) exposure ratings, Anthropic Economic Index, BLS OEWS May 2024. Not a job loss forecast. ### 05. Is your industry using AI? https://corelooppress.com/tools/is-your-industry-adopting/ Guess the share of firms in your industry using AI, then see the Census figure for your state and size. Source: U.S. Census Bureau Business Trends and Outlook Survey. ### 06. Who shows up in the compute data? https://corelooppress.com/tools/who-owns-the-compute/ Make seven guesses about AI chips, data centers and models, then see who appears in the partial public data. Source: Epoch AI open datasets (chip owners, GPU clusters, AI data centers, AI models), CC BY 4.0. Partial coverage, stated on the page. ### 07. How much power do data centers use? https://corelooppress.com/tools/power-bill-of-intelligence/ Make four estimates about U.S. and global data-center electricity, a 2030 projection and Virginia commercial-sales growth. Results compare your estimates with Berkeley Lab, IEA and EIA figures. The 2030 figure is a projection; Virginia commercial sales are not measured data-center consumption. ### 08. What do companies say about AI? https://corelooppress.com/tools/say-versus-spend/ Start with the fixed IC 100 company list across five groups, compare all 100 companies, then inspect filing passages, AI mentions, disclosure contexts, revenue, cash flow and total capex for a selected company. A separate, deeper filing-language study at https://corelooppress.com/tools/say-versus-spend/filing-language/ traces four deterministic text-pattern topics by fiscal year and group with linked SEC passages. Matches can miss synonyms or include hypothetical language; they do not establish activity or AI-only revenue. Total capex is not AI-specific spending. Roster and rule: https://corelooppress.com/tools/say-versus-spend/ic100/ ### 09. What's happening at the frontier? https://corelooppress.com/tools/say-versus-spend/frontier-ic25/ Explore 25 source-linked profiles across model builders, scale enablers and AI-native deployers. A dated Epoch Capabilities Index view compares selected builder models, not organizations. The fixed editorial roster is not a top-25 performance ranking or complete census. ### 10. What does a data center actually need? https://corelooppress.com/tools/data-center-resources/ Test six claims about servers, water, electricity and grid connections, then share an evidence map. Sources: IEA 2025 and 2026, Berkeley Lab 2024. Six claims; choices stay in the browser. ### 11. Compare country electricity systems https://corelooppress.com/tools/country-electricity/ Compare two of six places in 2022 or 2025: demand, generation, net imports and power mix. Source: Ember Yearly Electricity Data, September 2026 snapshot. A separate U.S. data-center overlay comes from Berkeley Lab. National totals do not establish local grid access. ### 12. How open is this AI? https://corelooppress.com/tools/open-or-closed/ Answer eight short scenarios about free APIs, downloadable weights, release licenses, data information, reproducibility, hosting cost and privacy controls. Every reveal links to a primary definition or provider document. Results show which of four distinctions the reader recognized and offer a downloadable share image. No quiz result classifies a model or a company. Answers stay in the browser; only aggregate interaction counts are sent. [Markdown guide and sources](https://corelooppress.com/tools/open-or-closed/index.md).